IEBGENER, Help

anandkumarrn · New User Joined: 06 Feb 2009 Posts: 23 Location: India

Hi,

I am trying to Concatenate 3 datasets to one dataset using IEBGENER, But i am getting error as mentioned below,

My lrecl and Blksize are all same for all the 4 datasets, can some one suggest me what would be the problem is?

DATA SET UTILITY - GENERATE

IEB351I I/O ERROR ,HDRTLNDM,HDRTRL ,9BE0,D, ,READ ,NO RECORD FOUND,00001B2000006,BSAM

MBabu · Posted: Sun Aug 09, 2009 12:10 am

are you able to BROWSE (not Edit or View) the data sets and scroll all the way to the bottom? What is the JCL?

anandkumarrn · New User Joined: 06 Feb 2009 Posts: 23 Location: India

dick scherrer · Posted: Sun Aug 09, 2009 1:12 am

Hello,

Is file2 or file3 empty? Is there a proper end-of-file? Simply allocating the file may not be sufficient. . .

What happens if you run the process with one file at a time?

MBabu · Posted: Sun Aug 09, 2009 7:13 am

I tried this and it works with my test files. Are these files Fixed or Variable? I'm wondering if a variable record might have a zero length. I don't know how to create a file like that so I can't test the idea.

If this is a 1 time task, you can try ISPF option 3.3 (move/copy) and select MOD on the 2nd panel

btw, there is no SYSOUT DD for the utilities

Bill Dennis · Posted: Mon Aug 10, 2009 8:43 pm

Is DSN=YYY actually used as both INPUT and OUTPUT?

anandkumarrn · New User Joined: 06 Feb 2009 Posts: 23 Location: India

MBabu · Posted: Mon Aug 10, 2009 9:54 pm

Bill, your eyes are better than mine! Depending on the amount of data in XXX and blocksizes, I'd expect it to maybe get an x37 abend as it just copies data it has already written over and over, but if it reads past the nonexistent EOF of YYY this error would make sense.

dick scherrer · Posted: Mon Aug 10, 2009 10:15 pm

Hello,

Robert Sample · Posted: Mon Aug 10, 2009 10:44 pm

If you really need to do this -- and most of the time it turns out not to be a requirement once looked at -- you will need to have a two-step job. The first step copies your three input files to a temporary file passed to the next step. The second step writes the temporary file back into the output file YYY as you called it.

In general, it is not a good idea to use the same file for input and output in one step unless you write the program to close the file between input and output operations -- for IEBGENER as the other utilities, you're not writing the program so you cannot control the processing. Therefore you should not use the same file for input and output for any utility.

Unpredictable results will occur when the same file is used for input and output at the same time -- often abends, sometimes corrupted data files.

Terry Heinze · Posted: Tue Aug 11, 2009 2:49 am

Given the DISP of SYSUT2, I'm surprised the job ran at all. Wouldn't there have been an exclusive enque put on it to prevent the execution?

MBabu · Posted: Tue Aug 11, 2009 7:03 am

Terry Heinze · Posted: Tue Aug 11, 2009 7:07 am

My mistake. I think I was thinking the job consisted of 2 steps. I think in a 2 step job the DISP=OLD in step 2 would have prevented the 1st step from even starting wouldn't it? I can't test it since I have no access to a mainframe.

Robert Sample · Posted: Tue Aug 11, 2009 4:18 pm

In JES3, the DISP=OLD would force exclusive access to the data set for the entire job.